Roundup Tracker - Issues

Message3948

Author stefan
Recipients rawler, richard, schlatterbeck, stefan
Date 2010-01-03.14:25:05
Message-id <4B40A8BD.4050607@codesourcery.com>
In-reply-to <1262520430.04.0.585352219529.issue2550514@psf.upfronthosting.co.za>
On 01/03/2010 07:07 AM, Ralf Schlatterbeck wrote:
>
> Ralf Schlatterbeck<rsc@runtux.com>  added the comment:
>
> When iterating over *many* results from a database query, we should have
> an iterator interface instead of fetching (possibly thousands of) all
> whole nodes at once. I've experimented with queries that return
> thousands of nodes and simply enlarged roundups node cache. The result
> is only marginally faster due to two things
> - lots of sql queries
> - memory allocation for the cache
>
> only the first will be addressed by the proposed patch...
> Maybe we return an iterator from getnodes?

I agree that an iterator is a good idea. Though the bottleneck we 
observed (and which let me to propose a change in the API) is due to IPC 
overhead caused by many individual queries, instead of few fused ones.
History
Date User Action Args
2010-01-03 14:25:06stefansetrecipients: + stefan, richard, schlatterbeck, rawler
2010-01-03 14:25:06stefanlinkissue2550514 messages
2010-01-03 14:25:05stefancreate